Electro thermal simulation of superconducting

advertisement
Supplementary Information
Francesco Marsili1,*, Faraz Najafi1, Charles Herder1, Karl K. Berggren1,2
1Department
of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge,
Massachusetts 02139, USA.
2Kavli
Institute of Nanoscience, Delft University of Technology, Lorentzweg 1, 2628CJ Delft, The Netherlands.
A. Parameters of the electro-thermal simulation.
We present and discuss additional parameters of the electro-thermal simulations in the following (see Table SI - I). The
substrate temperature used in the model was Tsub = 4.7 K, which was based on the temperature measured by a Si diode sensor
glued with cryogenic varnish to a detector chip mounted on our cryogenic-device-measurement setup 1. The nanowires
simulated in this paper were 30 nm wide and 4.5 nm thick, like the devices in ref. 1. The nanowire critical temperature was
TC = 10.8 K, which was measured on bare NbN films. The nanowire critical current was IC = 7.2 µA, as we measured on
30-nm-wide-nanowire SNSPDs 1. The coherence length of our NbN films was assumed to be ξ = 7 nm following ref. 2, which
reports on similar films. The nanowire inductance per square was L□ = 80 pH/□, which was estimated from the fall time of the
photoresponse pulse of 30-nm-wide-nanowire SNSPDs. The nanowire resistance per square was R□ = 680 Ω/□, which was
estimated from the ohmic branch of the I - V curves of 30-nm-wide-nanowire SNSPDs measured at 4.7 K.
Table SI - I. Parameters of the electro-thermal simulation.
Symbol
Quantity
Value
Tsub
substrate temperature
4.7 K
TC
critical temperature
10.8 K
IC
critical current
7.2 µA
ξ
Ginzburg-Landau coherence length
7 nm
L□
kinetic inductance per square
80 pH/□
R□
resistance per square
680 Ω/□
Tn
normal-slab temperature
A
Temperature coefficient of the
8.5 K
300 W/(m2K4)
heat-transfer coefficient per unit area
B. Derivation of the bolometric heat equation
The general form of the one-dimensional two-temperature heat equations for NbN is 3:
1
Ce Te  Te
t
Cp  Tp  Tp
t


Ce Te 
 e-p Te 
Ce  Te 
 e-p  Te 
T  T   j  
2
e
T
e
p
 Tp  
Cp  Tp 
 esc
 T Te 
 e 
e
x 
x 

T
 Tsub  
p
(1)
 
Tp 
 p  Tp 

x 
x 
(2)
where Te and Tp are the electron and phonon temperatures; Ce is the electron specific heat: Ce  Te in the normal state and
Ce  exp   Te  / kBTe  in the superconducting state 2; Cp  Tp3 is the phonon specific heat 2;  e  p  Te1.6 is the electronphonon interaction time 4; τesc is the phonon escape time to the substrate; κe and κp are the temperature-dependent electron and
phonon thermal conductivities; ρ is the NbN resistivity; j is the nanowire current density.
The bolometric heat equation is obtained adding equation (1) to equation (2) and setting Te = Tp = Tr:

t
C Tr  Tr  
j 
2
Cp Tr 
 esc
Tr  T  
sub
 T Tr 
 r 
x 
x 

(3)
where Tr is the reduced temeperature, C = Ce + Cp and κ = κe + κp.
We further simplified equation (3) into:

t
C Tr  Tr  
j 
2

d
Tr  T    Tr 
sub
 Tr
e
2
x
(4)
2
neglecting the phonon thermal conductivity and the spatial dependence of κe (∂κe / ∂x ∙ ∂Tr / ∂x ~ 0) as in ref. 5. We note that
we expressed the phonon-substrate coupling term in equation (3) in terms of the heat-transfer coefficient per unit area
α = Α ∙ Tr3 and the film thickness d. For the value and temperature dependence of C we followed ref.
2
and we estimated κe
from the nanowire resistivity as in 6.
C. Comparison between the bolometric and the 2-T heat equations
In this section we evaluate the validity of our choice of using the bolometric model over the 2-T model to describe the thermal
response of our nanowires.
The value of A / d that we estimated from the experimental value of the hotspot current was of the same order of magnitude,
but lower than the ratio between the temperature coefficient of Cp (Cp0) and τesc reported in ref.
2
for similar films:
A / d = 67 W/(mm3K4) ~ Cp0 / τesc = 125 W/(mm3K4). The fact that the phonon escape time estimated from our value of A / d
(146 ps) was a factor ~ 2 larger than in ref.
2
(78 ps) implied that we under-estimated the cooling of the phonons by the
2
substrate, which in fact partially compensated for over-estimating the cooling of the electrons by the phonons (for assuming
thermal equilibrium between the two sub-systems).
To quantitatively support this last claim, we compared the thermal response of a superconducting nanowire to an optical
excitation pulse simulated with the bolometric equation and our value of A, with the result of the 2-T equations relying on the
parameters reported in 2. For simplicity, we assumed the temperature to be homogeneous along the nanowire and then
neglected the thermal conduction terms. The excitation pulse had a peak optical power density of 1.5 mW / µm3 and a duration
of 300 ps, which reproduced the joule heating produced by a current of 3 µA flowing through a photon-induced normal
domain. The time evolution of the reduced temperature (Tr) simulated with the bolometric model was in agreement with the
average temperature (Tavg) between the electron and phonon temperatures simulated with the 2-T model (Figure SI - 1).
Considering that the results of the 2-T model were obtained with no free parameters, we concluded that our bolometric model
described the nanowire thermal response with an acceptable approximation respect to the more complete model.
Figure SI - 1. Simulated time evolution of the reduced temperature (Tr, in black), the electron temperature (Te, in red), the phonon temperature (Tp, in blue), and
the average between Te and Tp (Tavg, in blue).
1
F. Marsili, F. Najafi, E. Dauler, X. Hu, M. Csete, R. Molnar, and K. Berggren, submitted, available at: http://arxiv.org/ (2010).
2
A. D. Semenov, G. N. Gol'tsman, and A. A. Korneev, Physica C 351 (4), 349-356 (2001).
3
A. D. Semenov, G. N. Gol'tsman, and R. Sobolewski, Supercond. Sci. Technol. 15 (4), R1-R16 (2002).
4
Yu P. Gousev, G. N. Gol'tsman, A. D. Semenov, E. M. Gershenzon, R. S. Nebosis, M. A. Heusinger, and K. F. Renk, J. Appl. Phys. 75 (7), 36953697 (1994).
5
M. Tinkham, J. U. Free, C. N. Lau, and N. Markovic, Phys. Rev. B 68 (13), 134515 (2003).
6
J. K. W. Yang, A. J. Kerman, E. A. Dauler, V. Anant, K. M. Rosfjord, and K. K. Berggren, IEEE Trans. Appl. Supercond. 17 (2), 581-585 (2007).
3
Download